Engineering Geobacillus thermoglucosidasius for direct utilisation of holocellulose from wheat straw

Abstract

Background: A consolidated bioprocessing (CBP), where lignocellulose is converted into the desired product(s) in a single fermentative step without the addition of expensive degradative enzymes, represents the ideal solution of renewable routes to chemicals and fuels. Members of the genus Geobacillus are able to grow at elevated temperatures and are able to utilise a wide range of oligosaccharides derived from lignocellulose. This makes them ideally suited to the development of CBP.

Results: In this study, we engineered Geobacillus thermoglucosidasius NCIMB 11955 to utilise lignocellulosic biomass, in the form of nitric acid/ammonia treated wheat straw to which expensive hydrolytic enzymes had not been added. Two different strains, BZ9 and BZ10, were generated by integrating the cglT (β-1,4-glucosidase) gene from Thermoanaerobacter brockii into the genome, and localising genes encoding different cellulolytic enzymes on autonomous plasmids. The plasmid of strain BZ10 carried a synthetic cellulosomal operon comprising the celA (Endoglucanase A) gene from Clostridium thermocellum and cel6B (Exoglucanase) from Thermobifida fusca; whereas, strain BZ9 contained a plasmid encoding the celA (multidomain cellulase) gene from Caldicellulosiruptor bescii. All of the genes were successfully expressed, and their encoded products secreted in a functionally active form, as evidenced by their detection in culture supernatants by Western blotting and enzymatic assay. In the case of the C. bescii CelA enzyme, this is one of the first times that the heterologous production of this multi-functional enzyme has been achieved in a heterologous host. Both strains (BZ9 and BZ10) exhibited improved growth on pre-treated wheat straw, achieving a higher final OD600 and producing greater numbers of viable cells. To demonstrate that cellulosic ethanol can be produced directly from lignocellulosic biomass by a single organism, we established our consortium of hydrolytic enzymes in a previously engineered ethanologenic G. thermoglucosidasius strain, LS242. We observed approximately twofold and 1.6-fold increase in ethanol production in the recombinant G. thermoglucosidasius equivalent to BZ9 and BZ10, respectively, compared to G. thermoglucosidasius LS242 strain at 24 h of growth.

Conclusion: We engineered G. thermoglucosidasius to utilise a real-world lignocellulosic biomass substrate and demonstrated that cellulosic ethanol can be produced directly from lignocellulosic biomass in one step. Direct conversion of biomass into desired products represents a new paradigm for CBP, offering the potential for carbon neutral, cost-effective production of sustainable chemicals and fuels.

Keywords: Biomass; Cellulases; Consolidated bioprocessing (CBP); Endo/exoglucanases; Geobacillus thermoglucosidasius; Glycoside hydrolases; β-Glucosidase.

Fig. 1

Engineering G. thermoglucosidasius NCIMB 11955 for utilisation of lignocellulosic biomass. a The endoglucanase (CtCelA/CbCelA) acts on the low-crystallinity part of the cellulose fibre and create free chain-ends. The exoglucanases (Cel6B/CbCelA) then degrade the sugar chain by removing cellobiose units (dimers of glucose) from the free chain-ends. The released cellobiose units are finally hydrolyzed by β-glucosidases (CglT), releasing glucose. b The schematic illustrations of two strains, BZ9 and BZ10 containing chromosomally integrated cglT (β-1,4-glucosidase) gene from Thermoanaerobacter brockii. Strain BZ9 has a synthetic cellulosomal operon comprising CtcelA (Endoglucanase A) gene from Clostridium thermocellum, Ribosome binding site (RBS), Signal peptide (SP) and cel6B (Exoglucanase) from Thermobifida fusca heterologously expressed on a replicating plasmid having two origin of replications for Gram-negative (−ve rep) and Gram-positive (+ve Rep) bacteria under the constitutive strong promoter P_ldh. Strain BZ10 contains CbcelA (cellulase comprises a glycoside hydrolase family 9 and a family 48 catalytic domain) gene from Caldicellulosiruptor bescii on a replicating plasmid

Fig. 2

Expression of heterologous GHs by recombinant G. thermoglucosidasius strains. The concentrated supernatants from wild type as well as recombinant G. thermoglucosidasius strains expressing GHs either CtCelA or Cel6B or CbCelA were separated in SDS-PAGE and extracellular heterologous GHs at the indicated time points (2–9 h) were detected by Western blotting using ANTI-FLAG M2 monoclonal antibody-horseradish peroxidise conjugate. WT represents concentrated supernatant from wild-type strain of G. thermoglucosidasius at 8 h; M represents the pre-stained protein ladder (10–250 kDa). Expected molecular mass of the proteins are indicated by red arrows. a Congo red-stained TSA-CMC agar plate streaked with recombinant G. thermoglucosidasius expressing CtCelA enzyme (zone C1, C2, C3) and wild-type strain (zone WT). b Western blot showing the 100-fold concentrated supernatants from recombinant strain expressing the 50 kDa CtCelA. c CMCase specific activity of extracellular fraction of recombinant G. thermoglucosidasius expressing CtCelA and wild-type G. thermoglucosidasius strains (WT). d Western blot showing the 100-fold concentrated supernatants from recombinant G. thermoglucosidasius expressing the 60 kDa Cel6B. e RACase specific activity of extracellular fraction of recombinant G. thermoglucosidasius expressing Cel6B and G. thermoglucosidasius wild-type (WT) strains. f Western blot showing the 200-fold concentrated supernatants of recombinant G. thermoglucosidasius expressing the 193 kDa CbCelA. g CMCase and h RACase specific activity of extracellular fraction of recombinant G. thermoglucosidasius CbCelA and wild-type G. thermoglucosidasius strains (WT). P values were calculated by student’s t test and results are shown as mean ± SEM of three biological replicates

Fig. 3

Intracellular expression of CglT enhanced cellobiose utilisation. a Western blot showing FLAG-tagged CglT using ANTI-FLAG M2 monoclonal antibody-horseradish peroxidise conjugate. Lane Wt. and CglT represents soluble fraction of cell lysate from wild-type G. thermoglucosidasius and the recombinant G. thermoglucosidasius ZB3bInt strain, respectively. Lane M is pre-stained protein ladder (10–250 kDa); Expected molecular mass of 50 kDa for CglT is indicated by red arrow. b pNPGase specific activity of recombinant G. thermoglucosidasius ZB3bInt strains (ZB3bInt) and wild-type G. thermoglucosidasius strains (WT). P values were calculated by student’s t test and results are shown as mean ± SEM of three biological replicates. c Cellobiose consumption and growth profiles by recombinant G. thermoglucosidasius ZB3bInt (ZB3bInt) and wild-type G. thermoglucosidasius strains (WT) when grown on 3.0% cellobiose as the carbon source at 55 °C. Black circles, G. thermoglucosidasius_ZB3bInt (OD600 nm); Red circles, wild-type (OD600 nm); blue square, G. thermoglucosidasius_Zb3bInt (remaining cellobiose); green squares, wild-type (remaining cellobiose). Results are shown as mean ± SEM of three biological replicates

Fig. 4

Growth of engineered strains on pre-treated wheat straw. Colony forming units (CFU) were measured after growing recombinant G. thermoglucosidasius strains on pre-treated wheat straw for 12 h. a CFU comparison of recombinant G. thermoglucosidasius expressing only CbCelA (strain ZB6d) with G. thermoglucosidasius BZ9 strain expressing CglT and CbCelA. b CFU comparison of recombinant G. thermoglucosidasius BZ10 strain expressing CglT, Cel6B and CtCelA with G. thermoglucosidasius expressing only either CtCelA or Cel6B. P** ≤ 0.01, P*** ≤ 0.0001 were calculated by one-way ANOVA followed by Sidak’s multiple comparisons test. c Growth curve of recombinant G. thermoglucosidasius BZ9 and BZ10 strains on 1% pre-treated wheat straw as a sole carbon source, with wild-type G. thermoglucosidasius served as a control. Results are shown as mean ± SEM of three biological replicates

Fig. 5

Time course profile for production of ethanol from pre-treated wheat straw by recombinant G. thermoglucosidasius BZ243, BZ244 and LS242 strains. P**  ≤ 0.01, P**  ≤ 0.001, P***  ≤ 0.0001 were calculated by 2-way ANOVA followed by Tukey’s multiple comparisons test. Results are shown as mean ± SEM of three biological replicates